Mechanical Limit of lens opening steps

[donald_a]

I have an XTi and it allows changes in aperture setting in 1/3 stops, i.e.

f/5.6->f/6.3->f/7.1->f/8.0. My question is whether or not Canon lenses are

physically varying the aperture in such small steps? In other words if I set

my aperture to f/6.3 is the physical lens opening actually f/6.3 or is it f/5.6

with some compensating processing done in the camera?

[mark u]

<a target="_other" href="http://www.eosdoc.com/manuals?q=Av+steps" title="Av steps">Are the EOS apertures stepless?</a>

[mendel_leisk]

You can also get a feel for this by:

 

Shoot a static scene, with auto exposure (P mode). Note the f-stop and shutter speed, switch to Manual mode and set those, and shoot the same scene. More often than not you'll be able to see slight exposure difference, ie: the the scene darkens or brigthens a bit.

 

Wonder if something similar applies to shutter speeds?

[mark u]

See the link about shutter speeds in the link I provided...

[remco-jan.woldhuis]

I don't know for Canon but with a Nikon lens you can change the apertures stepless via the

level at the mount. I guess Canon has something similar.

[john_wilson10]

I don't know if they are stepless, but I believe they are changing for each 1/3rd stop. I don't know if your XTi has a depth of field preview button, but my 20D does, and when I hold it down while changing the aperture in manual mode, I can see the aperture stop down with every 1/3rd stop.

[andrew robertson]

Unlike with the Nikon, which has a mechanically actuated aperture, the EOS system has

electronic contacts only for all lens functions, so the actual precision is determined by

motors inside each lens. Heck, they may differ from each other internally.

[richard_oleson]

I'm sure they're stepless... operating them electronically from the body, steps would serve no function. They were only put in there in the first place to make it easier to set the position of the manual aperture ring.

[uncle_ziba]

Since the aperture is set by a step motor I'm not sure it can be completely "stepless".

 

You can easily check for yourself. Set the camera to manual or Av and hold

the DOF preview button. Now look at the aperture blades through front of the lens while changing the f stop. You would see that the blades are moving in 1/3 steps. from posts above it seems that in auto exposure modes the camera may use even smaller steps

[mark u]

I would take Canon's information that metering works in 1/8th stop increments internally to be correct (for a maximum "error" of 1/16th stop). That allows exposure calculations to be made with binary arithmetic. Since most EOS cameras can be adjusted to 1/6th stop (ISO in 1/3rd stop intervals on all but a few DSLRs together with half stop shutter speed/aperture in M mode again available on most bodies), it implies a slight disconnect at 1/6th, 1/3rd, 2/3rds and 5/6th stop. The disconnect is imperceptible to the human eye - 1/6-1/8=1/24th stop difference at most. In practice, mechanical repeatability can become a limiting factor, as anyone who has seen a lens or shutter with sticky aperture blades can verify. Note that the measured shutter speeds reflect an exact 1/8th stop step size.

[denny_wells]

This is an interesting topic that I'd never really considered - and Mark's link and discussion is certainly informative.

 

But what I'm left wondering is "does this matter"? Variations in printing processes, screen settings and viewing environments, light meter scene-evaluation accuracy (it still meters to grey, which you must remember when shooting a snowy landscape), and abilities to manipulate a RAW file for the XTi in question (or the film developing process for the more traditional SLR crowd) all dwarf the significance of 1/2 vs. 1/3. vs. 1/6 vs. 1/8 vs. continuous aperture precision.

 

Interesting technical stuff, but of no practical significance to my shooting.

[alan_rockwood]

I have essentially the same question about the rebel film cameras (Ti, T2, etc.), and also the same question with respect to shutter speeds.

 

I recently did a B&W film, and based on the look of the exposure/density curves it looks to me like the settings on a canon Ti are discontinuous, and in fact may be rather course-grained. (I will post a description of what I did if anyone is interested, but otherwise I won't use up forum space.)

[mark u]

If you look at the links and discussion you will find the answers - they are the same for film cameras and digital. You can repeat the experiments more easily yourself with a film camera, since you can shoot through the open back and observe. No film is needed for testing.